Project Summary Myxoma virus (MV) infects only rabbits in vivo, but also has a natural capacity to infect a wide variety of human cancer cells in vitro and in vivo. Thus, MV is an attractive candidate for oncolytic virotherapy to treat human cancer. MV has been used to successfully treat several diverse human brain cancers in xenografted immunodeficient mice and murine metastatic melanoma in immunocompetent mice. Here, MV will be developed for human clinical trials by exploiting a therapeutic strategy for which the virus is uniquely well- suited: ex vivo purging of cancer cells within human bone marrow or mobilized-PBMC samples from patients who would normally be excluded from autologous stem cell transplantation following high dose chemotherapy. Considerable preliminary data has been collected to support this proposal: 1) MV does not perturb or compromise human multipotent stem cell differentiation in immunodeficient mice engrafted with normal human bone marrow or cytokine-mobilized PBMCs, 2) MV eliminates a wide variety of human cancer cells following ex vivo purging, 3) MV can effectively purge not only permissive leukemia/lymphoma cells in vivo, but also unexpectedly prevents even nonpermissive human leukemia cells (such as KG1 cells) from engraftment or tumor induction, and 4) MV recombinants that express a variety of useful reporter proteins (fluorescent and bioluminescent) for imaging purposes have already been constructed. Specifically, our aims are: 1) Validate MV safety for ex vivo treatment of normal human hematopoietic stem cells: The safety of MV-purging for normal human stem cell differentiation will be tested, using engrafted immunodeficient NOG mice to verify full hematopoietic cell engraftment and immune reconstitution. MV purging will be tested on primary human stem/progenitor cells derived from normal bone marrow and G-CSF-mobilized PBMCs, using hematopoietic colony forming cell assays in vitro as well as for efficient hematologic cell engraftment in vivo. 2) Optimize MV ex vivo cancer cell purging: Two human cancers, B-cell lymphoma and acute myeloid leukemia, will be investigated for the ability of ex vivo MV purging to eliminate their tumorigenic potential in vivo in engrafted NOG mice. The cancer cells and viruses will be tagged with distinguishable luciferases that allow the engrafted tumor cells and the therapeutic virus to be independently tracked in vivo. Primary cells from acute myeloid leukemia patients will also be tested for the ability of MV to specifically eliminate the contaminating cancer cells and allow the selective engraftment of only noncancerous human leukocytes. 3) Investigate the mechanism of MV purging of primary human leukemia cells: We have recently shown that ex vivo infection of human KG1 leukemia cells with MV prevents the subsequent engraftment and tumor formation of these cells into NOG recipient mice, despite the fact that these cells are completely nonpermissive for MV infection in vitro. To assess for virus-induced cell signaling changes, we probed MV-infected KG1 cells with an array of antibodies to 46 different human signaling phosphoproteins, and observed that MV infection specifically induces Stat5 and Hck activation in KG1 cells. We will explore the functional significance of these host cell signaling activations for the successful ex vivo tumor cell purging of human leukemia calls by MV.